Pre-Grant Publication Number: 20100250597
Please help the USPTO examine the application by evaluating the relevance of the publicly submitted prior art to the patent application.
Peer To Patent forwards the Top 10 most relevant prior art submissions and their annotations to the USPTO.
Review this prior art and click on the thumbs up (or down) to indicate whether this submission should be forwarded to the USPTO.
If you login then you can add an annotation by typing in the box at the bottom of the screen to comment on the relevance of the prior art to the claims of the patent application.
Review this prior art and click on the thumbs up (or down) to indicate whether this submission should be forwarded to the USPTO.
If you login then you can add an annotation by typing in the box at the bottom of the screen to comment on the relevance of the prior art to the claims of the patent application.
Prior Art Detail
Summary / Description
| Summary / Description | Abstract: Concepts presented herein relate to extracting knowledge for a chatbot knowledge base from online discussion forms. Within a thread of an online discussion form, replies are selected based on structural features and content features therein. The replies can be ranked and used in a chatbot knowledge base. |
Basic Information
| Type of Prior Art | Issued Patents - US |
| Country | United States of America |
| Patent/Application # | US7814048 |
| Kind Code | United States (US) - United STATES Patent - A |
| Patentee Name | Microsoft Corporation |
| Relevant Pages, Columns, or Lines | Col. 1 Lines 20-25, Fig. 3 and |
| URL | http://www.freepatentsonline.co... |
| Filing Date | August 14, 2006 |
| Additional Information | |
Notes / To Do
| Notes | |
Excerpt
Excerpt Col. 2 Line 47 - Col. 3 Line 3:
FIG. 3 is a flow chart of a method 300 for extracting and ranking replies from a thread. At step 302, threads (i.e. thread 200) in an online discussion forum are accessed. At step 304, selected responses are identified within the threads based on structural features and content features of the replies.
It is desirable for the selected replies to be of high quality. The structural features and context features are used to identify quality responses. Structural features are indicative of a reply in a context of other replies in the thread. For example, the structural features can relate to whether the reply quotes the root message, quotes another reply, is posted by an author of the root message and the number of replies between the author's reply and a previous reply provided by the author.
Content features relate to words in a particular reply. For example, the features can include a number of words, a number of content words and/or a number of overlapping words. Content words are words that have some relationship to words in the root message and overlapping words are words that also appear in the root message. Additionally, the content features can relate to domain specific terms and whether the reply contains another person's nickname from the thread. Table 1 below lists example features that can be identified in step 304. These features are examples only and other features can also be used.
Col. 3 Lines 35 - 54:
FIG. 4 is a block diagram of a system 400 for extracting and ranking replies within a thread based on method 300 of FIG. 3. System 400 includes an identification module 402, a filter 404 and a ranking module 406. Identification module 402 receives a thread, for example thread 200. Identification module 402 identifies selected replies based on features of the reply. For example, identification module 402 identifies structural and content features as discussed above.
Filter 404 can filter out terms such as obscenities, personal information terms and/or forum specific terms within the responses identified by identification module 402. Additionally, replies that are duplicated and/or redundant can be removed from the selected replies. Ranking module 406 ranks the selected replies and generates a list 408 of (input, response) pairs. This list 408 is used in chatbot environment 100 of FIG. 1. For example, pairs can be stored in chatbot knowledge base 114 wherein the inputs can be used by pattern matching module 108 to match patterns in user input 104 and the responses can be used by response generator 112 to provide response 106.
Col. 4 Lines 49 - 56 discloses the applicability of these concepts to other computing environments beyond a chatbot.
|
Relevance
Claims
1
Relevance
A similar method for mining discussion threads is disclosed.
A similar method for mining discussion threads is disclosed.
Claim Chart
Some
2
Relevance
See the Excerpt, Col. 2 Line 47 - Col. 3 Line 3 and Col. 3 Lines 35-54.
See the Excerpt, Col. 2 Line 47 - Col. 3 Line 3 and Col. 3 Lines 35-54.
Claim Chart
Some
3
Relevance
See Col. 3 Lines 35-45:
Filter 404 can filter out terms such as obscenities, personal information terms and/or forum specific terms within the responses identified by identification module 402.
See Col. 3 Lines 35-45:
Filter 404 can filter out terms such as obscenities, personal information terms and/or forum specific terms within the responses identified by identification module 402.
Claim Chart
Some
0 days left
